Damage to visual cortex leads to direct impairment of vision with significant consequence to everyday life. Homonymous hemianopia, a loss of visual function on one side of space, is an extremely common and debilitating condition, yet it remains poorly understood and there are no effective treatments currently available. Despite this, considerable evidence exists to suggest stimuli presented to a blind hemifield can undergo processing and influence behaviour in a phenomenon known as ‘blindsight’ [1]. This can occur despite the individual often being completely unaware of the stimulus. The full nature and extent of such processing remains unclear, but it suggests residual processing within damaged cortex or recruitment of alternate visual pathways may occur. Here we evaluate the literature including important recent advances, and discuss how such residual processing may serve as a potential substrate for rehabilitation.

Clinical problem

Every year there are approximately 150,000 new cases of stroke in the UK, with at least a sixth estimated to have persistent hemianopia [2]. Similar deficits may follow traumatic brain injury and neoplasia. Hemianopia is notoriously difficult to treat, with spontaneous recovery unlikely after six months. Therapeutic targets to date include attempts to restore the deficit itself (the focus of this review), enlarge the field of gaze through compensatory mechanisms including saccadic eye-movements, and use of orthotic devices to increase the angle of vision in intact fields. In the majority of cases, such techniques have very limited benefit.

Residual processing

Past research hinted at great promise for the recovery of visual cortical disorders. With time, however, early positive findings have been questioned and the current situation is less clear. One critical issue is that the physiological target for rehabilitation has been unclear, in particular for complete hemianopia. We suggest that to make real progress, we must understand whether any residual processing takes place within a scotoma or blind hemifield. Such processing could then form a substrate for rehabilitative interventions.


In support of this notion, there is now substantial evidence for unconscious processing for stimuli presented in the blind field of hemianopes. One group of patients in particular has been investigated extensively, and may provide us with specific insights. ‘Blindsight’ is a term first used by Weiskrantz in the 1970’s in association with research on residual visual function after occipital cortex damage. The syndrome was defined following work on primates and the discovery that unlike humans, monkeys with lesions to primary visual cortex (V1 – see Figure 1) were still able to respond to visual events [1]. Some interpreted this as an evolutionary distinction between species, but it inspired research to match experimental scenarios (as best as possible) by demanding forced-choice responses from human subjects. This revealed – somewhat paradoxically – that some humans with blindness associated with V1 damage could also make a wide range of forced-choice visual discriminations including whether an object was present/absent, or coloured red/green, despite a complete lack of acknowledged awareness.

Figure 1: Visual cortex consists of V1-V5/MT with V1 (striate cortex) situated in the posterior pole of the occipital lobe. Shown are computationally inflated anatomical MRI scans (where gyri are light gray and sulci dark gray) of the occipital lobe on which have been superimposed functional MRI data, coloured to represent the retinotopic region of the visual field to which these areas respond. This is used to delineate the borders of retinotopic visual cortices, shown by dotted lines and labelled (the foveal representation is indicated by a star). Extrastriate retinotopic areas comprise dorsal (‘d’) and ventral (‘v’) regions representing the lower and upper visual fields respectively. Damage to the inferior division of the middle cerebral artery supplying lateral parietal and superior temporal cortices, as well as posterior cerebral artery damage supplying inferior temporal and occipital lobe, can result in prominent homonymous hemianopia.

Extensive work was undertaken subsequently to characterise responses in patients with ‘blindsight’ to stimuli distinguished by their motion, colour, and contrast, with the degree of subjective awareness varying considerably across these different manipulations [3]. It was initially suggested that residual processing might be due to surviving ‘islands’ of V1 that could process the stimuli. However, there is now substantial evidence refuting this claim [4]. Using a broad range of imaging and neurophysiology techniques to investigate the correlates of psychophysical findings, there is now evidence for activity in ipsilesional extrastriate cortex for stimuli presented within the ‘blind’ hemifield [5,6,7]. In particular, activation of the dorsal ‘where’ pathway (V5/MT) is common, often accompanied by a vague awareness of motion (at least via forced-choice) but little or no appreciation for the spatial structure of stimuli.

More recent work has measured responses in the ventral visual pathways of patients with blind-sight using images of natural, stationary objects such as faces or fruit. In one study, event-related potentials gave a typical positive occipital deflection (‘P1 response’) in ipsilesional occipital cortex for stimuli presented in the scotoma [8]. Similar targets were used in functional MRI (fMRI) experiments [9] comparing ventral with dorsal stimuli (rotating spirals), and activity in extrastriate cortex was found despite the patients having no awareness for any of the targets (see Figure 2). However the full extent of such processing remains unknown, including whether cortical responses are non-specific, or depend upon the category or features of individual objects.

Figure 2: Brain activation data for blindsight patient GY, presented with coloured images of natural objects in either sighted or blind hemifields. i. fMRI results superimposed on a subpart of the flattened cortical representation of each hemisphere. Dark red regions correspond to the lesion. Activated regions responding solely to objects in the left hemifield are coloured green, with regions responding to the ‘blind’ right hemifield in red. Areas responding with equal strength to stimuli in either hemifield are shown in yellow. In GY, the normal hemisphere (RH) responds to stimuli in the normal left hemifield with a similar activation pattern to controls. When stimuli were presented in the right hemifield, there was also response seen in ventral areas of the lesioned left hemishere (LH), including a region in the fusiform gyrus (FG) and lateral occipital region. ii. These graphs illustrate time course in different cortical areas for left and right hemifield stimulation (green and red curves respectively); location of plotted areas are indicated by numbers 1-4. Adapted from Goebel et al with permission [9].

Concerns

Most of the studies reviewed above investigated only one or two patients, often recruiting the same ‘blindsight’ subject, GY. At times he exhibited blind-sight [13], at others not, and on occasions it was claimed that he showed conscious awareness of stimuli in the hemianopic field [5]. This could be explained if such situations depend upon saliency [10a]. It was acknowledged for some time that stimuli particularly high in contrast or speed may be more effective at evoking sensations that blindsight patients (or others interacting with them) might interpret as visual consciousness – even if the visual experience cannot be described in any detail.

More recently, it has been found that stimuli with specific spatio-temporal properties are particularly effective at inducing blindsight in patients with hemianopia. Using stimuli optimised in this way, it appears that ‘blindsight’ may exist in a large proportion of patients with hemianopia, suggesting blindsight is more common than first imagined [11]. Clinically this is interesting because after documenting hemianopia, clinicians rarely attempt to elicit forced-choice responses for stimuli presented in the blind hemifield. Blindsight may therefore not be a small subgroup but instead might represent the majority of hemianopes.

Plasticity and Pathways

It is important to establish the route and mechanism for residual visual processing following V1 damage. There has been much speculation whether information may be transmitted subcortically via superior colliculus and pulvinar (the retinotectal route), or directly from thalamus to extrastriate cortex. Several groups used the temporal resolution afforded by electrophysiology attempt to infer which anatomical route reflected residual processing by studying the timing of evoked responses to visual stimuli, but results were extremely variable [6,7]. While one study suggested extrastriate activity occurred early relative to V1 activity, another suggested that extrastriate activity followed normal V1 activity, suggesting a slower alternate route.

| Sign up to the ACNR email newsletter here |

Recent work in non-human primates with chronic V1 lesions shows a clear causal role for the LGN and pathways from LGN to extrastriate cortex in V1-independent processing of visual information [12]. Monkeys were able to detect targets presented within a scotoma caused by V1 damage, and this was associated with contralateral extrastriate cortical responses. However, chemical inactivation of the LGN led to abolition of virtually all these extrastriate responses. This suggests that direct projections from LGN to extrastriate cortex support residual vision.

However, such findings do not explain whether activity observed in human extrastriate cortex following V1 damage comes from intact residual pathways, or the innate ability of the human cortex to reorder and organise following injury. Bridge and colleagues [13] recently utilised diffusion-weighted MRI tractography to investigate structural plasticity in the visual cortex of a patient with chronic V1 damage (Figure 3). They showed that normal pathways bypassing V1 from LGN to extrastriate cortex (MT+/V5) remain intact. They also demonstrated two new connections not seen in healthy individuals, from contralesional LGN to ipsilesional extrastriate cortex, and extrastriate cortico-cortical connections bilaterally. This certainly suggests plasticity may be involved following visual cortex damage through rearrangement of long-range connections, perhaps driven by a compensatory reaction to brain damage. Such findings in humans do not necessarily contradict the findings in non-human primates, if information is required from LGN bilaterally to generate blindsight [22].

Figure 3: Cortico-cortical pathways in control subjects and GY, who has left V1 destroyed. The top row shows the connection through the splenium between MT+/V5 bilaterally, and the lower row shows the connection between V1 bilaterally. The strength of the pathways from the centre of the splenium to MT+ /V5 bilaterally in GY is suggestive of an increase in cortico-cortical connectivity between this area in the two hemispheres. There is a 16-fold increase in the strength of the connection when the tract is seeded from the left and runs to the right. The size of the connection from unlesioned right hemisphere to left MT+/V5 is 10 times greater than the mean for the controls. Adapted from Bridge et al., with permission [13].

Rehabilitation

Visual rehabilitation for hemianopia has often proved unsuccessful and current treatment strategies are limited. One recent development is ‘visual restoration therapy’ (VRT), a simple computer-based programme for patients to use for an hour a day (Figure 4) [14,15]. Despite early suggestions that VRT benefited vision following hemianopia, attempts at replication with more rigorous methods did not support earlier findings [16]. One concern is that saccadic eye movement was not fully controlled for, nor appropriately taken into account. This was clearly demonstrated in a subsequent study by using a scanning laser ophthalmoscope, considered one of the most effective ways to control eye movement as experimenters visualise fixation on the fundus throughout the task, and discard trials whenever saccades are made. Using such eye movement control, whether VRT was efficacious was questioned and other criticisms have also arisen [17]. One consideration was that it did not take into account the extensive research on unconscious processing in hemianopia.

Figure 4: Example set-up for visual rehabilitation therapy. The patient fixates upon the central fixation point, whilst being presented with successive stimuli within the blind (right) hemifield. For example in Sahraie et al. [18] stimuli would be presented for 2 seconds to three different regions of the receptive field per session, and 50 trials in each. After each trial, the patient reports the interval containing the grating patch by pressing one of two response buttons. This task was carried out everyday over a 3-month period.

More recently it has been suggested that hemianopia can be rehabilitated by employing stimuli custom-designed to undergo processing by the pathways mediating ‘blindsight’ or residual vision. By using salient targets with particular spatial and temporal properties in a rehabilitation tool, Sahraie and colleagues demonstrated improvement in sensitivity and awareness deep within visual field defects in 12 patients trained for three months [18]. They also suggest that feedback accelerates the effect [10b]. This approach seems promising, and represents one of the first approaches that can generate improvement deep within the scotoma. Nevertheless these early findings require replication and extension.

Moreover, some of the criticisms of other rehabilitative approaches may also be relevant. Sahraie’s study also did not control eye movements, although they argue that there is now evidence that recovery is independent of eye movement strategies. The previous group responded to concerns by applying quantitative analysis to eye movements before and after VRT. They found no change, suggesting instead that the time-consuming microperimetry was too difficult for subjects to accurately engage in [19].

Future

The visual system is the major sensory system in humans, yet it is still not clear whether it has the same capacity for recovery as, for example, the motor cortex. There remains significant disparity between studies, contributed to by differences in paradigms, stimuli, as well as measures of outcome. Recently, however, there is a trend towards consistent reports, and the use of blindsight as a way of defining potentially spared pathways that could be targeted by novel rehabilitative approaches is promising.


One approach is to systematically explore unconscious processing in the ventral visual pathway of healthy volunteers, and to generalise this to patients with brain damage causing hemianopia. Although beyond the scope of this article, the findings in heminaopia are strikingly similar to recent evidence in healthy volunteers that substantial unconscious processing takes place when visual stimuli are rendered invisible through masking or other paradigms [see 20 for a detailed review]. Also, one can look further at structural plasticity, and whether functional changes following rehabilitation [18,21] are reflected in changes of connectivity.

Conclusion

Residual visual processing can undoubtedly occur following brain injury in homonymous hemianopia. Recently, evidence has started to emerge to suggest a viable pathway for this, and that plasticity in the brain may alter these connections following brain injury. By understanding the nature and extent of unconscious processing, as well as how such changes may occur within the visual system, we hope to achieve a reliable target for neurorehabilitation of hemianopia – an extremely common and debilitating condition. 

References

1. Weiskrantz L Roots of blindsight. Prog Brain Res. 2004: 144: 229-41.

2. Townend BS, Sturm JW, Petsoglou C, O’Leary B, Whyte S, &Crimmins D Perimetric homonymous visual field loss post-stroke. J Clin Neurosci 2007: 14(8):754-6.

3. Stoerig P, & Cowey A Wavelength sensitivity in blindsight. Nature 1989: 342 (6252): 916-918.

4. Stoerig P, Kleinschmidt A, & Frahm J No visual responses in denervated V1: high-resolution functional magnetic resonance imaging of a blindsight patient. Neuroreport 1998: 9(1):21-5.

5. Barbur JL, Watson JD, Frackowiak RS, & Zeki S Conscious visual perception without V1. Brain 1993: 116(6):1293-302.

6. ffytche DH, Guy CN, & Zeki S Motion specific responses from a blind hemifield. Brain 1996: 119(6): 1971-82.

7. Holliday IE, Anderson SJ, & Harding GF Magnetoencephalographic evidence for non-geniculostriate visual input to human cortical area V5. Neuropsychologia 1997: 35(8): 1139-46.

8. Rossion B, de Gelder B, Pourtois G, Guérit JM, & Weiskrantz L Early extrastriate activity without primary visual cortex in humans. Neurosci Lett. 2000: 279(1):25-8.

9. Goebel R, Mucklib L, Zanellac FE, Singerb W, & Stoerig P Sustained extrastriate cortical activation without visual awareness revealed by fMRI studies of hemianopic patients. Vis Res 2001: 41: 1459-74.

10a. Sahraie A, Hibbard PB, Trevethan CT, Ritchie KL, Weiskrantz L Consciousness of the first order in blindsight. Proc Natl Acad Sci U S A. 2010: Nov 15. Ahead of print.

10b. Sahraie A, Macleod MJ, Trevethan CT, Robson SE, Olson JA, Callaghan P, & Yip B Improved detection following Neuro- Eye Therapy in patients with post-geniculate brain damage. Exp Brain Res. 2010: 206(1):25-34.

11. Sahraie A, Trevethan CT, & MacLeod MJ Temporal properties of spatial channel of processing in hemianopia. Neuropsychologia 2008: 46: 879-85.

12. Schmid MC, Mrowka SW, Turchi J, Saunders RC, Wilke M, Peters AJ, Ye FQ, & Leopold DA Blindsight depends on the lateral geniculate nucleus. Nature 2010: 466: 373-7.

13. Bridge H, Thomas O, Jbabdi S, & Cowey A Changes in connectivity after visual cortical brain damage underlie altered visual function. Brain 2008: 131: 1433-1444.

14. Zihl J, & von Cramon D Visual field recovery from scotoma in patients with postgeniculate damage. A review of 55 cases. Brain 1985: 108(2):335-65.

15. Kasten E, Wüst S, Behrens-Baumann W, & Sabel BA Computer-based training for the treatment of partial blindness. Nat Med. 1998: 4(9):1083-7.

16. Reinhard J, Schreiber A, Schiefer U, Kasten E, Sabel BA, Kenkel S, Vonthein R, & Trauzettel-Klosinski S Does visual restitution training change absolute homonymous visual field defects? A fundus controlled study. Br J Ophthalmol. 2005: 89(1):30-5.

17. Horton JC Vision restoration therapy: confounded by eye movements. Br J Ophthalmol. 2005: 89(7):792-4.

18. Sahraie A, Trevethan CT, MacLeod MJ, Murray AD, Olson JA,& Weiskrantz L Increased sensitivity after repeated stimulation of residual spatial channels in blindsight. PNAS 2006:103: 14971-14976.

19. Kasten E, Bunzenthal U, & Sabel BA Visual field recovery after vision restoration therapy (VRT) is independent of eye movements: an eye tracker study. Behav Brain Res. 2006: 175(1):18-26.

20. Lin Z, & He S Seeing the invisible: The scope and limits of unconscious processing in binocular rivalry. Progress in Neurobiology: 2009: 87: 195–211.

21. Nelles G, de Greiff A, Pscherer A, Forsting M, Gerhard H, Edger J, & Diener HC Cortical activation in hemianopia after stroke. Neurosci Lett. 2007: 426(1):34-8.

22. Silvanto J, Cowey A, Lavie N & Walsh V Making the blind-sighted see. Neuropsychologia. 2007: 45(14):3346-50.